Automatic Number Plate Detection using Deep Learning

Abstract

Vehicle Number plate detection is a technology which uses Convolutional Neural Network to detect and recognize the number plate data. The objective is to design an efficient automatic number plate detector using the vehicle images. The system can be implemented for security purposes in the restricted areas like military zones, government offices, parking management and toll management. The developed system first applies various image pre-processing techniques using OpenCV. Vehicle number region is extracted using the image segmentation. The trained neural network detects and recognizes the characters in the number plate. The resulting data can be stored in the database and can compare with the record on a database. The system performance is tested on real image. It is noticed from the testing that the established system effectively detects and recognize the vehicle number plate on factual images.